⚡️ Speed up function process_description by 6%

[codeflash-ai]

📄 6% (0.06x) speedup for process_description in pr_agent/algo/utils.py

⏱️ Runtime : 4.16 milliseconds → 3.94 milliseconds (best of 582 runs)

📝 Explanation and details

The optimized code achieves a 5% speedup through strategic regex pattern pre-compilation and loop optimization:

Key Optimizations:

1. Pre-compiled Regex Patterns: All regex patterns are now compiled once at module load time instead of being recompiled on every function call. The most impactful change is the global _file_walkthrough_pattern that's compiled only once and reused, eliminating the expensive re.split() with re.DOTALL flag compilation overhead.

2. Efficient Pattern Matching Loop: Instead of trying multiple regex patterns sequentially with separate re.search() calls, the code now loops through pre-compiled patterns in _fallback_patterns and breaks on the first match. This reduces regex execution time, especially when the first pattern matches.

3. Reduced Regex Compilation Overhead: The line profiler shows the original code spent significant time in re.split() (5% of total time) and multiple re.search() calls (13.7% + 7.8% + 8.1% = 29.6% combined). The optimized version reduces this by eliminating repeated pattern compilation.

Performance Characteristics:

• Best for large-scale cases: The optimizations show the most improvement on test cases with many files (10-15% faster) where regex compilation overhead compounds
• Consistent gains: Even simple cases show 1-3% improvements due to reduced setup overhead
• Memory efficient: Pre-compiled patterns are shared across all function calls rather than recreated each time

The optimizations preserve all original functionality while eliminating redundant regex compilation work that becomes more expensive as the input size grows.

✅ Correctness verification report:

Test	Status
⚙️ Existing Unit Tests	🔘 None Found
🌀 Generated Regression Tests	✅ 33 Passed
⏪ Replay Tests	🔘 None Found
🔎 Concolic Coverage Tests	🔘 None Found
📊 Tests Coverage	82.6%

🌀 Generated Regression Tests and Runtime

from typing import List, Tuple

# imports
import pytest  # used for our unit tests
from pr_agent.algo.utils import process_description


class PRDescriptionHeader:
    FILE_WALKTHROUGH = type("EnumValue", (), {"value": "Files Changed Walkthrough"})()
from pr_agent.algo.utils import process_description

# unit tests

# -------- BASIC TEST CASES --------

def test_empty_string_returns_empty_tuple():
    # Empty input should return empty string and empty list
    base, files = process_description("") # 341ns -> 328ns (3.96% faster)

def test_no_walkthrough_header_returns_full_description():
    # If the walkthrough header is missing, return the full description and empty files
    desc = "This is a PR description without walkthrough."
    base, files = process_description(desc) # 1.13μs -> 1.17μs (3.33% slower)

def test_simple_walkthrough_header_split():
    # Basic split on header, no HTML tags
    desc = "Intro text.\nFiles Changed Walkthrough\nSome details here."
    base, files = process_description(desc) # 1.13μs -> 1.16μs (2.41% slower)

def test_html_header_split_and_file_extraction():
    # Basic HTML structure with one file walkthrough
    desc = (
        "Intro text.\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>full/path/foo.py<ul>* Change 1<br> * Change 2</ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.22μs -> 1.20μs (2.00% faster)

def test_multiple_files_extraction():
    # Multiple files in the walkthrough section
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary1</code></summary><hr>foo/path/foo.py<ul>* Foo change</ul></details>"
        "</td></tr>"
        "<tr><td>"
        "<details><summary><strong>bar.py</strong><dd><code>Summary2</code></summary><hr>bar/path/bar.py<ul>* Bar change</ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.24μs -> 1.26μs (1.43% slower)

# -------- EDGE TEST CASES --------

def test_walkthrough_header_but_no_details():
    # Walkthrough header present but no details section
    desc = "Header\nFiles Changed Walkthrough\n"
    base, files = process_description(desc) # 1.07μs -> 1.09μs (1.93% slower)

def test_walkthrough_header_and_details_but_no_files():
    # Walkthrough header and HTML details but no <tr><td> blocks
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.11μs -> 1.09μs (1.46% faster)

def test_walkthrough_with_nonstandard_html():
    # Walkthrough header with malformed HTML (missing closing tags)
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>* Change 1"
        # missing closing tags
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.15μs -> 1.15μs (0.087% faster)

def test_walkthrough_with_extra_text_after_table():
    # Walkthrough header with extra text after the table
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>* Change 1</ul></details>"
        "</td></tr>"
        "</table>\n\n___\nSome extra text"
    )
    base, files = process_description(desc) # 1.19μs -> 1.16μs (3.29% faster)

def test_walkthrough_with_no_table_end_marker():
    # Walkthrough header with no table end marker
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>* Change 1</ul></details>"
        "</td></tr>"
        "</table>"
    )
    base, files = process_description(desc) # 1.08μs -> 1.13μs (4.41% slower)

def test_walkthrough_file_with_hyphen_bullet():
    # Walkthrough file summary with hyphen instead of bullet
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>- Change 1</ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.16μs -> 1.07μs (8.01% faster)

def test_walkthrough_file_with_code_ellipsis():
    # File with <code>...</code> should be skipped
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>...</code></summary><hr>foo/path/foo.py<ul>* Change 1</ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.16μs -> 1.15μs (1.31% faster)

def test_walkthrough_with_unusual_whitespace():
    # Walkthrough with extra whitespace and line breaks
    desc = (
        "Header\n"
        "<details>   <summary>   <h3>Files Changed Walkthrough</h3>   </summary>"
        "<table>\n"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong>   <dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>\n* Change 1\n</ul></details>"
        "</td></tr>\n"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.17μs -> 1.14μs (3.44% faster)

def test_walkthrough_with_br_tags_in_summary():
    # Walkthrough with <br> tags in summary
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>* Change 1<br> * Change 2<br></ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.14μs -> 1.15μs (1.13% slower)

# -------- LARGE SCALE TEST CASES --------

def test_large_number_of_files():
    # Test with 500 files in the walkthrough
    file_template = (
        "<tr><td>"
        "<details><summary><strong>{name}</strong><dd><code>{summary}</code></summary><hr>{full_name}<ul>* {change}</ul></details>"
        "</td></tr>"
    )
    table_rows = "".join(
        file_template.format(
            name=f"file_{i}.py",
            summary=f"Summary {i}",
            full_name=f"path/to/file_{i}.py",
            change=f"Change {i}"
        )
        for i in range(500)
    )
    desc = (
        "Bulk update\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        f"{table_rows}"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 20.8μs -> 20.8μs (0.038% slower)
    for i in range(500):
        pass

def test_large_description_text():
    # Test with a very large description before walkthrough
    large_text = "A" * 10000
    desc = (
        f"{large_text}\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        "<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>* Change 1</ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.69μs -> 1.66μs (1.68% faster)

def test_large_file_summaries():
    # Test with large summaries in file details
    large_summary = "Change " + "X" * 500
    desc = (
        "Header\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        "<tr><td>"
        f"<details><summary><strong>foo.py</strong><dd><code>Summary</code></summary><hr>foo/path/foo.py<ul>* {large_summary}</ul></details>"
        "</td></tr>"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 1.16μs -> 1.16μs (0.086% slower)

def test_walkthrough_with_just_under_1000_files():
    # Test with 999 files (max scale for guidelines)
    file_template = (
        "<tr><td>"
        "<details><summary><strong>{name}</strong><dd><code>{summary}</code></summary><hr>{full_name}<ul>* {change}</ul></details>"
        "</td></tr>"
    )
    table_rows = "".join(
        file_template.format(
            name=f"file_{i}.py",
            summary=f"Summary {i}",
            full_name=f"path/to/file_{i}.py",
            change=f"Change {i}"
        )
        for i in range(999)
    )
    desc = (
        "Big update\n"
        "<details><summary><h3>Files Changed Walkthrough</h3></summary>"
        "<table>"
        f"{table_rows}"
        "</table>\n\n___"
    )
    base, files = process_description(desc) # 39.7μs -> 39.7μs (0.055% faster)
# codeflash_output is used to check that the output of the original code is the same as that of the optimized code.
#------------------------------------------------
from typing import List, Tuple

# imports
import pytest  # used for our unit tests
from pr_agent.algo.utils import process_description


class PRDescriptionHeader:
    FILE_WALKTHROUGH = type("Enum", (), {"value": "File Walkthrough"})()
from pr_agent.algo.utils import process_description

# unit tests

# Basic Test Cases
def test_empty_description():
    # Should handle empty string input
    desc, files = process_description("") # 324ns -> 352ns (7.95% slower)

def test_no_file_walkthrough():
    # Should return the description and empty files if no walkthrough header
    input_str = "This is a PR description without walkthrough."
    desc, files = process_description(input_str) # 1.05μs -> 1.11μs (4.70% slower)

def test_simple_file_walkthrough_split():
    # Should split on the walkthrough header and return base + empty files if no details
    base = "Base description."
    walkthrough = "File Walkthrough"
    input_str = f"{base}{walkthrough}"
    desc, files = process_description(input_str) # 88.8μs -> 85.4μs (4.02% faster)

def test_html_walkthrough_single_file():
    # Should parse a single file walkthrough entry correctly
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py<ul>•Implemented new function</details>'
        '</td></tr></table>\n\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 42.1μs -> 38.1μs (10.5% faster)

def test_html_walkthrough_multiple_files():
    # Should parse multiple file entries
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table>'
        '<tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py<ul>•Implemented new function</details>'
        '</td></tr>'
        '<tr><td>'
        '<details><summary><strong>file2.py</strong><dd><code>Bug fix</code></summary><hr>src/file2.py<ul>•Fixed bug</details>'
        '</td></tr>'
        '</table>\n\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 45.7μs -> 41.4μs (10.4% faster)

# Edge Test Cases

def test_walkthrough_with_unusual_html():
    # Should handle missing <ul> and different bullet symbols
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py-Implemented new function</details>'
        '</td></tr></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 39.2μs -> 34.8μs (12.7% faster)

def test_walkthrough_with_missing_details():
    # Should skip files that cannot be parsed
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>...</code></summary><hr>src/file1.py<ul>•Could not analyze</details>'
        '</td></tr></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 31.7μs -> 27.8μs (14.1% faster)

def test_walkthrough_with_extra_whitespace():
    # Should handle extra whitespace and newlines gracefully
    base = "Base description."
    walkthrough_html = (
        '   <details>   <summary>   <h3>File Walkthrough</h3>   </summary>   '
        '<table>   <tr>   <td>   '
        '<details>   <summary>   <strong>file1.py</strong>   <dd>   <code>Added function</code>   </summary>   <hr>   src/file1.py<ul>•Implemented new function   </details>   '
        '</td>   </tr>   </table>\n___   '
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 96.6μs -> 93.2μs (3.68% faster)

def test_walkthrough_with_no_table_end_marker():
    # Should handle missing end markers gracefully
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py<ul>•Implemented new function</details>'
        '</td></tr></table>'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 33.5μs -> 29.2μs (14.5% faster)

def test_walkthrough_with_no_files():
    # Should handle valid walkthrough header but no file entries
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 12.9μs -> 9.67μs (33.9% faster)

def test_walkthrough_with_html_entities():
    # Should decode HTML entities in summaries
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added &amp; improved</code></summary><hr>src/file1.py<ul>•Implemented &lt;b&gt;bold&lt;/b&gt; text</details>'
        '</td></tr></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 59.7μs -> 56.3μs (6.08% faster)

# Large Scale Test Cases

def test_large_number_of_files():
    # Should handle up to 500 files efficiently
    base = "Base description."
    file_details = ""
    for i in range(500):
        file_details += (
            f'<tr><td>'
            f'<details><summary><strong>file{i}.py</strong><dd><code>Summary {i}</code></summary><hr>src/file{i}.py<ul>•Change {i}</details>'
            f'</td></tr>'
        )
    walkthrough_html = f'<details><summary><h3>File Walkthrough</h3></summary><table>{file_details}</table>\n___'
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 3.49ms -> 3.32ms (5.11% faster)
    for i in range(0, 500, 50):  # Check every 50th file for correctness
        pass

def test_large_description_string():
    # Should handle very large base description and a single file
    base = "A" * 1000
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py<ul>•Implemented new function</details>'
        '</td></tr></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 35.8μs -> 31.1μs (15.1% faster)

def test_large_file_summary():
    # Should handle very large summary in a file entry
    base = "Base description."
    large_summary = "Change " + ("A" * 900)
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        f'<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py<ul>•{large_summary}</details>'
        '</td></tr></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 68.4μs -> 64.4μs (6.36% faster)

def test_large_file_walkthrough_header():
    # Should handle many walkthrough headers (should split only on first)
    base = "Base description."
    walkthrough_html = (
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file1.py</strong><dd><code>Added function</code></summary><hr>src/file1.py<ul>•Implemented new function</details>'
        '</td></tr></table>\n___'
        'Some text'
        '<details><summary><h3>File Walkthrough</h3></summary>'
        '<table><tr><td>'
        '<details><summary><strong>file2.py</strong><dd><code>Bug fix</code></summary><hr>src/file2.py<ul>•Fixed bug</details>'
        '</td></tr></table>\n___'
    )
    input_str = f"{base}{walkthrough_html}"
    desc, files = process_description(input_str) # 31.1μs -> 27.4μs (13.6% faster)
# codeflash_output is used to check that the output of the original code is the same as that of the optimized code.

To edit these changes git checkout codeflash/optimize-process_description-mgzg90au and push.